Determination of the myoglobin states in ground beef using non-invasive reflectance spectrometry and multivariate regression analysis.

Seventy-two samples of ground beef from M. semimembranosus of two 5 and two 1.5year old animals were prepared. Two types of fat tissues from either beef or pork were added to the ground beef. The samples were prepared to contain predominantly deoxymyoglobin (DMb), oxymyoglobin (OMb) and metmyoglobin (MMb) states on surfaces using selected methods based on chemical treatment (for MMb) and oxygen pressure packaging to induce the two other states. Reflectance spectra were measured on ground beef after three storage times. Partial least regression analysis was used to make calibration models of the desired myoglobin states. Validated models using leave-one-sample out cross validation gave, after correction and normalization, prediction errors of about 5%. Long term storage of ground beef was unsuitable for preparing pure MMb states due to gradual reduction of the pigment to DMb, presumably by bacteria.

New procedure for improving precision and accuracy of instrumental color measurements of beef.

The surface layers of steaks from bovine M. semimembranosus were prepared to have deoxy- (DMb), oxy- (OMb) and metmyoglobin (MMb) states using either chemicals (CHEM) or oxygen partial pressure packaging (OPP). Ninety-six different meat surface areas were measured in reflectance mode (400-1100 nm) for each preparation method. Reflectance spectra were converted to absorbance (A) and then transformed by Kubelka-Munk transformation (K/S) and/or extended multiplicative scatter correction (EMSC). Transformed spectra of prepared pure states were used to make calibration models of MMb, DMb and OMb using either selected wavelengths (SW) or partial least square (PLS) regression. Finally, the predicted myoglobin states were normalized to ensure that no state was <0 or >1 and the sum of all states equal to 1. Multivariate calibrations (i.e. PLS) outperformed the univariate calibrations (i.e. SW). The OPP method of preparing pure states was clearly best for OMb while the CHEM method was best for preparing MMb on fresh meat surfaces. Both preparation methods needed improvement concerning DMb. The CHEM(K/S) SW and the OPP EMSC(A) PLS methods predicted MMb, DMb and OMb with root-mean-square errors of cross validation (RMSECV) equal to 0.08, 0.16 and 0.18 (range 0-1) and 0.04, 0.04 and 0.04 (range 0-1), respectively. This new reflectance protocol has potential for routine meat color measurements.

Scattering correction by use of a priori information.

In this paper we demonstrate how a limited amount of a priori knowledge about spectral variability can be used in extended multiplicative scattering correction (EMSC) to remove disturbing effects such as light scattering variation in visible and near-infrared spectra prior to data modeling. Two different datasets were studied. In the first dataset, pigment concentrations (astaxanthin) were estimated in a model system with different concentrations of the scattering agent intralipid. Different cases were created by including varying levels of intralipid in the calibration set and then applying the models on sample sets with scattering properties both within and outside the calibration range. Including the most accurate estimate of light scattering in the EMSC model gave root mean square errors of prediction (RMSEP) that were similar to a cross-validated global model including all samples, even for extreme extrapolation with regard to scattering properties. Less accurate estimates gave on average RMSEPs half of what could be achieved using EMSC without any a priori knowledge, suggesting that the method also has potential in cases where the accurate light scattering spectrum is difficult to obtain. In the second dataset carbohydrate concentrations (sucrose, fructose, and glucose) were estimated in orange-juice mixtures where unwanted spectral variation was caused by a change in distance between transmittance fiber-optic probes. This caused two different interfering phenomena due to path length variation and saturation in the detection system. The prediction results for a model based on spectra collected at one specific probe distance treated with EMSC with a correction spectrum were comparable to what could be achieved by a global model including spectra collected at three different distances. The corresponding RMSEPs for models using EMSC with no correction term were in the worst cases 4, 22, and 36 times higher for sucrose, fructose, and glucose, respectively.

Long-term stability of a Raman instrument determining iodine value in pork adipose tissue.

The stability of a calibration model from a non-destructive Raman instrument during a period of three years was studied. A calibration model created on a dataset measuring pork adipose tissue in 2005 determining iodine value (IV), was transferred to a dataset measuring pork adipose tissue three years later in 2008. During these three years the fibre optic cable had been changed and the output of the laser was reduced to 60% compared with the power in 2005. The samples were also taken from different parts of the carcass. Aligning the peak positions and pre-processing with multiplicative scatter correction together with a selection of wavelengths/wavenumbers gave, for IV, a correlation coefficient of 0.95 for measured versus predicted IV of the 2008 samples. The accuracy expressed as root mean square error of prediction was 2.04 g iodine added to 100g of melted fat with 6 partial least squares factors for the 2008 samples. This study shows that it is possible, with minor modifications, to transfer the model from spectra measured three years later on the same instrument. It is concluded that a quantitative use of Raman instruments are robust over time.

Determination of omega-6 and omega-3 fatty acids in pork adipose tissue with nondestructive Raman and fourier transform infrared spectroscopy.

In order to predict omega-6 and omega-3 fatty acids in the diet of humans, seventy-three pork back fat adipose tissue samples were measured with Raman spectroscopy directly on adipose tissue and on melted fat. Melted fat samples were, in addition, measured with Fourier transform infrared (FT-IR) spectroscopy. Gas chromatography analyses were conducted as the reference analysis. Partial least squares regression (PLSR) was used to calibrate and validate all models predicting omega-3 and omega-6 fatty acids contents from spectra. Omega-6 fatty acids in melted fat measured with FT-IR was predicted with a correlation coefficient (R) of 0.93 and a root mean square error of cross-validation (RMSECV) of 1.61% of the total amount of fatty acids. Raman spectra measured on melted fat gave a prediction of omega-6 fatty acids with R=0.97, and RMSECV=0.99% of total amount of fatty acids. Omega-6 fatty acids were predicted with R=0.94, and RMSECV=1.50% of the total amount of fatty acids using Raman spectra recorded on adipose tissue. For omega-3 fatty acids, the highest R=0.91, and lowest RMSECV=0.23% of the total amount of fatty acids were obtained from Raman spectra acquired on melted fat. FT-IR and Raman spectroscopy may be used as rapid, nondestructive methods to determine omega-6 and omega-3 fatty acids in melted fat. Raman spectroscopy can also be used directly on adipose tissue.

Quantitative determination of saturated-, monounsaturated- and polyunsaturated fatty acids in pork adipose tissue with non-destructive Raman spectroscopy.

The composition of dietary fat has received increased attention during the recent years because it influences human health. Seventy seven samples from pork adipose tissue and melted fat from the same tissue were measured with Raman spectroscopy. Gas chromatography analysis was conducted as reference. Iodine values (IV) ranged from 58.2 to 90.4g iodine added per 100g fat. Polyunsaturated fatty acids (PUFA) ranged from 7.8% to 31.7% and monounsaturated fatty acids (MUFA) from 35.2% to 51.5% of total fatty acids. When applied on pre-processed spectra of melted fat, partial least square regression (PLSR) with cross-validation gave a correlation coefficient (R)=0.98, and root mean square error of cross-validation (RMSECV)=1.4 for IV, using 3 PLS factors in the model. PUFA gave R=0.98 and RMSECV=1.0% of total fatty acids, using 5 PLS factors. MUFA were predicted with R=0.96 and RMSECV=1.0% of total fatty acids, using 9 PLS factors. On adipose tissue a model with 3 PLS factors gave R=0.97 and RMSECV=1.8 for IV. For PUFA, a model with 3 PLS factors gave R=0.95 and RMSECV=1.5% of total fatty acids. For MUFA a model with 6 PLS factors gave R=0.91 and RMSECV=1.5% of total fatty acids. The results indicate the feasibility to use Raman spectroscopy as a rapid and non-destructive method to determine IV, PUFA, MUFA and saturated fatty acids (SFA) measured directly on pork adipose tissue and in melted fat from the same tissue.

Low-cost approaches to robust temperature compensation in near-infrared calibration and prediction situations.

The traditional way of handling temperature shifts and other perturbations in calibration situations is to incorporate the non-relevant spectral variation in the calibration set by measuring the samples at various conditions. The present paper proposes two low-cost approaches based on simulation and prior knowledge about the perturbations, and these are compared to traditional methods. The first approach is based on augmentation of the calibration matrix through adding simulated noise on the spectra. The second approach is a correction method that removes the non-relevant variation from new spectra. Neither method demands exact knowledge of the perturbation levels. Using the augmentation method it was found that a few, in this case four, selected samples run under different conditions gave approximately the same robustness as running all the calibration samples under different conditions. For the carbohydrate data set, all robustification methods investigated worked well, including the use of pure water spectra for temperature compensation. For the more complex meat data set, only the augmentation method gave comparable results to the full global model.

Comparison of methods for transfer of calibration models in near-infared spectroscopy: a case study based on correcting path length differences using fiber-optic transmittance probes in in-line near-infrared spectroscopy.

This article addresses problems related to transfer of calibration models due to variations in distance between the transmittance fiber-optic probes. The data have been generated using a mixture design and measured at five different probe distances. A number of techniques reported in the literature have been compared. These include multiplicative scatter correction (MSC), path length correction (PLC), finite impulse response (FIR), orthogonal signal correction (OSC), piecewise direct standardization (PDS), and robust calibration. The quality of the predictions was expressed in terms of root mean square error of prediction (RMSEP). Robust calibration gave good calibration transfer results, while the other methods did not give acceptable results.

Determination of C22:5 and C22:6 marine fatty acids in pork fat with Fourier transform mid-infrared spectroscopy.

Fatty acids in samples (n=74) of pork adipose tissue were measured with a Fourier transform mid-infrared (FT-MIR) spectrometer and by gas chromatography. The measured absorption spectra provided information to estimate partial least squares regression models for fatty acid groups, the iodine value and several fatty acids. The iodine values were predicted with correlation coefficient R=0.996 and root mean square error of cross-validation RMSECV=0.658. The sum of the two marine fatty acids of main interest, C22:5n3+C22:6n3, were predicted with R=0.982 and RMSECV=0.062. The K nearest neighbours procedure successfully classified the samples in three classes, depending on their proportions of marine fatty acids. Application of fat and absorption measurements were rapid, requiring less than 5 min of labour per sample. The results reported in this paper demonstrate that FT-MIR measurements can serve as a rapid method to determine marine fatty acids in pork fat.

Accurate wavelength measurements of a putative standard for near-infrared diffuse reflection spectrometry.

The diffuse reflection (DR) spectrum of a sample consisting of a mixture of rare earth oxides and talc was measured at 2 cm-1 resolution, using five different accessories installed on five different Fourier transform near-infrared (FT-NIR) spectrometers from four manufacturers. Peak positions for 37 peaks were determined using two peak-picking algorithms: center-of-mass and polynomial fitting. The wavenumber of the band center reported by either of these techniques was sensitive to the slope of the baseline, and so the baseline of the spectra was corrected using either a polynomial fit or conversion to the second derivative. Significantly different results were obtained with one combination of spectrometer and accessory than the others. Apparently, the beam path through the interferometer and DR accessory was different for this accessory than for any of the other measurements, causing a severe degradation of the resolution. Spectra measured on this instrument were removed as outliers. For measurements made on FT-NIR spectrometers, it is shown that it is important to check the resolution at which the spectrum has been measured using lines in the vibration-rotation spectrum of atmospheric water vapor and to specify the peak-picking and baseline-correction algorithms that are used to process the measured spectra. The variance between the results given by the four different methods of peak-picking and baseline correction was substantially larger than the variance between the remaining five measurements. Certain bands were found to be more suitable than others for use as wavelength standards. A band at 5943.13 cm-1 (1682.62 nm) was found to be the most stable band between the four methods and the six measurements. A band at 5177.04 cm-1 (1931.61 nm) has the highest precision between different measurements when polynomial baseline correction and polynomial peak-picking algorithms are used.